Handwriting input method and apparatus

Abstract

A handwriting input method and apparatus is disclosed which generates a single line segment from a group of input line segments which are input by a user. The input line segments may be overlapping or disconnected short line segments, thereby allowing the user to enter lines in a free stroke format, much the way an artist sketches. For each of the input line segments a plurality of variable parameters are detected from the stroke data. The parameters for each input line segment are converted to a three-dimensional weighting function. The weighting functions for different line segments are combined, the edge of the resulting combined function is detected, and that edge is displayed as a line segment on a display device.

We claim: 
     
       1. A method for receiving strokes handwritten on an input/output screen and displaying a line segment on said screen, said method comprising the steps of: receiving a stroke input by writing by hand with a pen on a first area of an input/output screen;   associating with each point of said stroke a plurality of parameters by detecting the pen-stroke pressure P of said stroke   detecting the pen velocity V of said stroke;   detecting the number of times N previous strokes have been input in said local area;   converting P, V and N to a plurality of parameters wherein the following parameters are calculated: certainty factor Cf as a parameter showing whether the user is writing a line segment, Cf being a function of at least P and V   merge factor Mf as a parameter showing the strength with which the user writes a line segment, Mf being a function of at least P;   joint factor Jf as a parameter showing density of the user input, Jf being a function of at least V; and   repetition factor Rf as a parameter showing whether the user has intention of repeatedly writing a line segment, Rf being a function of the number of previous strokes input into said local area;       weighting said plurality of parameters to generate a three-dimensional weighting function, said weighting function associated with said input stroke;   reading from a memory means a stored weighting function associated with a previously input stroke, said previous stroke input in a local area on said input/output screen, said local area being an area substantially close to said first area;   adding said weighting function associated with said input stroke and said stored weighting function, said result of said addition being a generated weighting function;   storing said generated weighting function in said memory means;   detecting the edge of said weighting function to extract a line segment; and   displaying said line segment on said input/output screen.   
     
     
       2. A method as set forth in claim 1 which further comprises: said steps of receiving stroke data, associating a plurality of parameters with said data, generating a weighting function, adding said weighting function with a previously stored weighting function, storing the result of said addition, and detecting the edge of said result, is performed iteratively. 
     
     
       3. A method as set forth in claim 1, wherein: Cf is calculated by a relation of Cf=kc×P/V where kc is a constant;   Mf is calculated by a relation of Mf=km×P where km is a constant;   Jf is calculated by a relation of Jf=kj×V where kj is a constant; and   Rf is calculated by a relation of Rf=kr×R(N) (kr is a constant).   
     
     
       4. A method as set forth in claim 1 in which: the pen-stroke pressure P and pen velocity V at is detected at each input point AN of said input stroke; and   said certainty factor Cf and merge factor Mf are calculated for said each input point AN as a function of said pen-stroke pressure P and pen velocity V at said input point;   joint factor Jf iS calculated upon completion of said stroke, Jf being associated with the initial and final ends of said input stroke as a function of said pen velocity V at all points AN of said input stroke;   repetition factor Rf is calculated by reading out the data corresponding to repetition factors for previously input strokes, said previously input stokes input in said same local area as said current stroke, Rf being calculated from said previous repetition factors by increasing the number of repetition factors by one; and   a three-dimensional weighting function Vf is calculated as a function of said calculated Cf, Mf, Jf and Rf.   
     
     
       5. A method as in claim 4 wherein: said three-dimensional weighting function Vf is converted to bit map data;   said bit map data is normalized and stored in said memory means;   said bit map data is converted to two-dimensional parameters by edge detection, thereby generating a line segment; and   said line segment is displayed on said screen.   
     
     
       6. A handwriting input apparatus comprising: an input/output means for receiving data from a plurality of strokes drawing by a user;   said input/output means including a detection means for extracting an input state of each of said plurality of strokes, said input states comprising a plurality of variable parameters;   a processing means including: a weighting function generating means for generating a three-dimensional weighting function on the basis of said plurality of variable parameters associated with each of said plurality of strokes, said plurality of variable parameters comprising parameters based on the stroke pressure P, velocity V, and coordinate values of each of said strokes, and number of repetitions N associated with each of said strokes, that is the number of said plurality of strokes which are drawn in a same local area;   wherein, for performing the conversion from said input state amounts P, V and N to three-dimensional parameters, said detecting means includes: a certainty detector for obtaining certainty factor Cf as a parameter showing whether the user is writing a line segment with certainty by a relation of Cf=kc×P/V (kc is a constant),   a merge degree detector for obtaining merge factor Mf as a parameter showing the strength with which the user writes a line Segment by a relation of Mf=km×P (km is a constant),   a joint degree detector for obtaining joint factor Jf as a parameter showing the density of the user input by a relation of Jf=kj×V (kj is a constant), and   an input area detector and repetition degree detector for obtaining repetition factor Rf as a parameter showing whether the user has intention of repeatedly drawing line segments by a relation of Rf=kr×R(N) (kr is a constant), and   in said weighting function generating means, said weighting function Vf is generated from said certainty factor Cf, joint factor Jf, merge factor Mf and repetition factor Rf;       a local area reading means for reading out said weighting function associated with each of said plurality of strokes input in a local area;   a bit map handling means for adding two or more of said weighting functions to generate a new weighting function;   a recording means to record said new weighting function in a memory means;   an edge detecting means for detecting the edge of said new weighting function, said edge associated with a line segment;   a display means to display said line segment.   
     
     
       7. A handwriting input apparatus as set forth in claim 6 in which said input states are detected for each point of said input stroke. 
     
     
       8. A handwriting input apparatus as set forth in claim 6 wherein Rf is determined from the number of times of repetition RP(N) that a coordinate point AN is associated with each of said plurality of strokes by converting RP(N) to Rf by means of a table.